Proceedings of the International Conference on Applied Science and Technology on Engineering Science 2025 (iCAST-ES 2025)

<Previous Article In Volume
Next Article In Volume>

LiteGator: A Lightweight and Budget-Aware Retriever Routing Framework for Low-Resource Question Answering

Authors
Ar-Razy Muhammad Darmanto1, *, Rizqia Lestika Atimi1
1Department of Electrical Engineering and Informatics, Politeknik Negeri Ketapang, Ketapang, Indonesia
*Corresponding author. Email: ar-razy.muhammad@politap.ac.id
Corresponding Author
Ar-Razy Muhammad Darmanto
Available Online 31 December 2025.
DOI
10.2991/978-94-6463-926-1_77How to use a DOI?
Keywords
Adaptive-retrieval system; Low-Resource retriever routing; Retrieval-augmented generation
Abstract

Retrieval-Augmented Generation (RAG) systems rely on efficient and accurate retrievers to supply relevant context to large language models. While dense retrievers achieve superior semantic recall, they are computationally expensive, limiting their practicality in latency-sensitive or resource-constrained environments. In this work, we introduce LiteGator, a lightweight and adaptive retriever routing framework that dynamically selects between sparse and dense retrieval based on shallow query features. Without requiring model fine-tuning or retriever modifications, LiteGator employs a linear Support Vector Machine (SVM) trained on a small query set to make routing decisions in milliseconds. Evaluated on three QA benchmarks—Natural Questions, TriviaQA, and Scifact QA—LiteGator achieves competitive retrieval quality while significantly reducing average latency compared to dense and hybrid baselines. Our results highlight the feasibility of retrieval strategies that are fast, interpretable, and budget-aware for real-world QA systems.

Copyright
© 2025 The Author(s)
Open Access
Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

<Previous Article In Volume
Next Article In Volume>
Volume Title
Proceedings of the International Conference on Applied Science and Technology on Engineering Science 2025 (iCAST-ES 2025)
Series
Advances in Engineering Research
Publication Date
31 December 2025
ISBN
978-94-6463-926-1
ISSN
2352-5401
DOI
10.2991/978-94-6463-926-1_77How to use a DOI?
Copyright
© 2025 The Author(s)
Open Access
Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Cite this article

risenwbib
TY  - CONF
AU  - Ar-Razy Muhammad Darmanto
AU  - Rizqia Lestika Atimi
PY  - 2025
DA  - 2025/12/31
TI  - LiteGator: A Lightweight and Budget-Aware Retriever Routing Framework for Low-Resource Question Answering
BT  - Proceedings of the International Conference on Applied Science and Technology on Engineering Science 2025 (iCAST-ES 2025)
PB  - Atlantis Press
SP  - 687
EP  - 695
SN  - 2352-5401
UR  - https://doi.org/10.2991/978-94-6463-926-1_77
DO  - 10.2991/978-94-6463-926-1_77
ID  - Darmanto2025
ER  -